Congestion pricing in Australia should be a no-brainer

Centrist Australian think-tank, the Grattan Institute, has released a report advocating congestion charging for Australian cities.  The report is a good summary of the fundamental problem of traffic congestion and the strategies adopted by Australian cities and states in addressing it.

None of this will be news to readers of this blog.  The key point being that it is almost impossible to sustainably address traffic congestion in major cities by simply building capacity (paid for largely by those not using that capacity) to meet demand, whether it be capacity on roads or on public transport (which is commonly seen as the main way to attract traffic off of roads).   It cites the avoidable costs of congestion from a BITRE study of (Bureau of Infrastructure, Transport and Regional

Economics) of A$6.1 billion in Sydney and A$4.6 billion in Melbourne.  This is a figure imputed from the costs of lost travel time (and vehicle operating costs), but is still an economic drain.  There is no plausible way of significantly reducing these costs without pricing to spread and moderate demand.

I think the report provides a quite compelling case for congestion pricing in Sydney and Melbourne.  It particularly includes research and data that is pertinent to other "new world cities", characterised by largely low density population and land use, high private car ownership and usage and dispersed employment locations.  Many assumptions that may be widely held among decision makers and the public should be challenged by this report.

Some of the highlights are the following:

• In the morning peak up to 21 per cent of trips on Sydney roads are for socialising, recreation, or shopping (i.e. not commuting, or trips to education) (p.8).  This infers that the scope to price some of those trips onto other modes or at other times should be significant, and more importantly, even a drop of a quarter of those trips would likely have a noticeable effect in reducing congestion).  (The figure for Melbourne is 11%).  It also might infer that the elasticity of demand for those trips in the morning peak is greater than for others, but this ought to be established by further research;
• There is record spending on urban road and public transport infrastructure in major cities (over A$35 billion in the current year), indicating that it isn't a lack of spending on supply that is the issue (p.12), and the majority of committed spending is on public transport (p.8).  Quite simply, building more capacity will never be enough (and the value of that spending continues to drop);
• ANPR technology is now the most feasible option to use for cordon and corridor charging (p.13), as toll tags are increasingly unnecessary;
• There is insufficient use of "repurposing road space", which can be used to increase overall capacity or provide dedicated capacity to specific road users.  On average, 14% of road space can be reallocated (typically to cycling and pedestrians) without reducing overall capacity (p.22);
• Parking levies have very limited impact (A$2490 for Sydney CBD, A$1440 for Melbourne), noting that up to 40% of vehicles in the Sydney CBD are through traffic (compared to a third in Melbourne). (p.25);
• CBD cordons in Sydney/Melbourne could improve speeds by up to 16% in the CBDs and 20% on roads approaching them, and 1% improvement in whole of network speeds (pp.28-29), with more details to come in a report next week;
• CBD cordons would mostly affect high income drivers, as it is them who predominantly drive to the CBDs.  Only 15% of jobs in Sydney and Melbourne are in the CBD (pp.35-36);
• People on higher incomes tend to drive the furthest to work, 30% of workers live in the suburb they work in, or an adjacent one (p.36). Which may also indicate that charging by distance will mostly affect those on higher incomes;
• Low income drivers with few alternatives can be protected from excessive impacts of congestion charging (p.40).
• The report claims "now is the time" because others are doing it, but this shouldn't be the only determinant.  Of the proposals listed, Hong Kong is on hold for fairly obvious reasons, Vancouver's proposals received a very poor public response and are unlikely to proceed, Jakarta's proposals have been fraught with a range of difficulties (which I have written about on this blog).  I doubt in the short term whether any US city, other than New York, will advance further given the politics and lack of creative policy thinking (p.10).

The report rightfully (and in contrast to some other reports lately) notes there are broadly three main options for charging:

1.   Cordons (this should include area charging), although it only talks about CBD (central city) cordons, when this tool could be applied more widely onto other centres of activity.  London (as an area charge), Stockholm, Gothenburg, Milan, Valetta and the future New York and Abu Dhabi schemes are all cordons, and Singapore has one as part of its scheme;

2.   Corridor charges, although again this could be wider than a major highway and could include charges on viable alternative routes. Singapore and Dubai both have corridor charges; and

3.  Network charges, which it defines only as distance based charging, but actually needs to disaggregate by route and time of day (simply charging all travel at a flat rate by distance within an area wouldn't achieve much in comparison).  No city has this for congestion pricing to date, although Singapore will be implementing the technology that could facilitate this in the next year.

The reaction

Sadly I'm not surprised that the political reaction has been poor.  With the possible exception of former (Federal) Minister for Urban Infrastructure Paul Fletcher, there is at best a void of interest in congestion pricing in Australia and at worst antipathy which demonstrates fear most of all.

Of course, the experience of London is well known, and there was a flurry of interest in the UK in the five years after London implemented its congestion charge, but other schemes came to nought, for a range of reasons including lack of trust that charging elsewhere could deliver improvements for those paying that were worthwhile, and antipathy towards yet another increase in the cost of motoring. 

However, things have changed elsewhere.  The United States, where car use is dominant in all cities (except lower Manhattan), now has a flurry of interest in investigating congestion pricing. New York is proceeding, but the jury is out on other cities. Closer to Australia, work has continued on congestion pricing in Auckland (indeed Auckland has had multiple studies on congestion pricing specifically or considering pricing as part of a wider package for over 15 years).  The fact that new world cities are seriously considering it ought to mean the same should happen in Australia.

However in Australia the reaction from most circles is a big fat no, which is exactly what came from several sources in the days after the report was released.  Victorian Premier Daniel Andrews (who was re-elected in 2018 with an increased majority), who has a strong reputation for action on climate change said (according to the ABC):

The best way to ease congestion is to build a public transport network system which can deliver more trains, more often — and we're getting it done....We have no plans and do not support a congestion tax.

New South Wales Premier Gladys Berejiklian (who was re-elected earlier this year) said pretty much the same (according to 7News):

The best way to reduce congestion into the future is to build major public transport projects

The NSW Transport Minister echoed this. 

Unfortunately, they are all wrong.

London and Paris have public transport networks that would be the envy of any Australian city, but the simple rule is that large cities cannot build themselves out of congestion with public transport or roads, if pricing is not used as a tool to manage demand.

It is almost a cliché to say "building new roads just generates more demand", but this is in a climate of not applying efficient pricing to that capacity. 

But what about the toll roads?

Ah but Sydney and Melbourne have toll roads you say.  Yes they do, but only some major roads are tolled and almost none of them have higher prices at peak times (the Sydney Harbour crossings do, but the difference between peak and off peak prices are so small (A$1) as to have a correspondingly small impact).

Road Australia map of Sydney toll roads including those under construction

The negative for Sydney, Melbourne and Brisbane is that residents of those cities are highly likely to see congestion pricing as "just another toll", and media coverage of the issue reinforces this.  Because some toll roads are regularly congested, there is likely to be a high degree of scepticism that congestion pricing at modest levels would reduce congestion, when relatively high tolls do not appear to have that effect (but of course they DO have the effect of reducing demand on those roads, but as long as they remain priced the same all day long, there wont be any real difference in demand patterns compared to untolled roads, except parallel routes in very low traffic volume periods). Furthermore, as many toll roads are private concessions with concession agreements that limit policy options to constrain the revenue from tolls for the concessionaires, practically speaking it could be difficult to implement congestion pricing on a wide scale without having to compensate investors in those toll roads.

Melbourne toll roads in red

In other words, tolling is a negative when it is unpopular and linked to a choice of using a new road which is tolled compared to an existing road.  Yes tolls in Sydney and Melbourne contribute to moderating demand, but that effect is not apparent because most toll roads have the same price all day long.

What should happen?

States should investigate congestion pricing as a tool to reduce traffic congestion, sustainably manage demand on the road networks, encourage mode and time of day travel shift (very few commentators really note that part of congestion pricing is changing when people drive not just how people travel.

Congestion pricing is obviously thought of as a way of generating more revenue to spend on transport, but it could also be used to replace or reduce existing charges. For example, registration fees could be cut state wide, benefiting those in regional and rural areas who have virtually no transport alternatives.  Private vehicle registration fees in Australian state are high compared to New Zealand and US states (e.g. Victoria charges up to A$834.80 a year). 

Furthermore, congestion pricing options should be developed based on making noticeable improvements to network performance NOT revenue raising, and the debate about recycling the revenue can proceed.

Be very clear, there will be severe traffic congestion in Sydney and Melbourne for many decades, no matter how much money politicians pour into roads and public transport.  It wont be significantly eased without the use of pricing.  It's about time that work was undertaken to investigate options, to engage with the public about such options, what they would mean in terms of winners and losers, and how congestion pricing could reduce the burden of registration fees for everyone (much better than the ludicrous NSW toll relief on registration fees).

What I predict is that Auckland will have congestion pricing by 2025, even on a small scale, but by then the debate wont have moved on in Australia at the political level, if the politicians themselves don't get investigations undertaken about congestion pricing.

Grattan Institute proposes Sydney and Melbourne congestion charge cordons

Across some of the Australian media is the release of a report from the Grattan Institute (a public policy thinktank) report titled "Stuck in traffic? Road congestion in Sydney and Melbourne" (PDF) by Marion Terrill.  It needs a rethink.

It reports what those of us working in the road pricing field would say is, generally speaking, fairly obvious.  Charging for road use at peak times in cities can reduce congestion and is worthwhile.  However, for all of the general merit of the argument, and some useful data in the report, I question very much one of the key conclusions of the report and the value of the report to the public debate about congestion.  In fact, I'd suggest it is counterproductive and sends the case for road pricing backwards in Australia.  It doesn't help that the report doesn't even understand the London scheme properly.

Let me be clear, I am not talking about parking pricing or public transport pricing, all of which have some merits.  Independent regulation of toll road pricing is interesting, but naive.  With private concessions already, there is effectively a contract with private concessionaires about pricing.  Overall the report is lacking in some fairly fundamental analysis, as it provides selected data that indicates on the one hand that most car trips in both cities are not to the CBD.  29% of trips to the CBD are by car in Melbourne, only 15% in Sydney.  81% of trips in greater Melbourne are by car, 67% in Sydney are by car.  Yet its solutions would indicate that this should be the focus.  

My problem with the Grattan Institute report is threefold:

1. The proposed solution of cordon charging for the CBDs of Melbourne and Sydney;
2. The lack of any reference to progress on reforms that could eventually lead to road pricing in cities;
3. The failure to emphasise that the main difficulty with the introduction of congestion pricing is public acceptability.

Why not cordons?

The report briefly mentions London, Stockholm and Singapore cases of congestion charging, but fails to acknowledge that London is an area charge and conditions in central London are now as slow as they were before the congestion charge was introduced in 2003.  Indeed the report completely misconstrues the London congestion charge as follows: 

But the gains in travel speeds are slowly diminishing, due to steadily growing traffic volumes and an inherent limitation of cordon schemes – vehicles that stay inside the zone are not charged, making it free for them to cruise the inner London streets.

The travel speeds are back to where they were before the charge was introduced, in part because road space has been reallocated to pedestrians, cyclists and bus lanes, but also because of uncharged vehicle growth (private hire vehicles - prebooked taxis).  Vehicles that stay in the zone ARE charged because London has an area charge, so they are not free to cruise the inner London streets.  Indeed a key part of the London problem is not that, but that almost half of all vehicles entering central London are either exempt or have a 100% discount from the congestion charge.  The Grattan Institute report ignores this.

None of these examples (and Gothenburg is a much less convincing example) have urban form similar to Sydney or Melbourne.  All have higher densities of population, all have urban commuting patterns more concentrated on their CBDs than the two biggest Australian cities.  In short, the car use patterns in Sydney and Melbourne are much more about people moving between suburbs and within them, than on long trips to the CBDs.  Cordons for Melbourne and Sydney could make a difference to those areas, but the impacts beyond the CBDs are likely to be relatively minor.  

If most car commuting in the major cities is not about going to the CBD, then charging trips to the latter are unlikely to make a big difference to most congestion.  Furthermore, the report dismisses the boundary effects of introducing a cordon charge.  What impact does it have on home or businesses on the "wrong" side having to pay a one off charge for a short trip?  Does it mean a cordon should be ruled out?  No.  However, the idea that this is the right solution is intellectually lazy.  

Much more likely to be effective would be network road pricing, which is what the Grattan Institute says but then doesn't recommend (it thinks that a cordon charge is network road pricing, but it certainly is not).  That means paying to use roads on a network wide basis, varying by time of day and location.  Obviously this would be a much bigger step than a cordon, but I am unsure why other road pricing options are ignored.  Furthermore, although it is acknowledged that such charges could offset registration fees (which seems odd in that it would mean commuters would get an offset of registration fees, but others wouldn't) and that net revenues should be spent on public transport.  Surely if it is offsetting registration fees it should be offsetting spending on roads?  Is there a case for more public transport spending per se or is it assumed?  Surely the idea that revenues should be spent based on merit would make more sense.

This comes to my second point.

What about road reform?

The Grattan Institute appears to be completely unaware of the national Heavy Vehicle Road Reform programme and the proposal in that to create an independent price regulator for existing and future road charges (existing being registration fees and fuel tax).  Heavy Vehicle Road Reform envisages a future whereby all heavy vehicles pay by mass, distance and location.  Furthermore, the Commonwealth government announced some months ago that there would be a study launched into road charging for light vehicles.  The latter, in part due to the challenges in the future as fuel efficiency, hybrid and electric vehicles erode fuel tax revenues.  The Grattan Institute seems oblivious to the likely introduction of an independent price regulator for national heavy vehicle charges, or the possible introduction of full network charges for heavy vehicles at least.   

You see congestion pricing should be seen in a wider context, in that how roads are managed and charged for should be reformed more fundamentally.  That means moving from fixed (registration) charges and fuel taxes to distance, mass, location and time of day charges, set by utility based road providers with an independent price regulator.  Some support for wider reform would have been helpful, but lack of acknowledgement of road reform seems odd.

Public acceptability?

Well this is the key problem, and the media coverage in Australia following the release of the report has almost entirely been negative.  Who believes that charging for road use will reduce congestion?  No one, and it is in part because the Sydney Harbour crossing peak charging has had negligible impacts, but moreso because it hasn't been piloted seriously in Australia.  There is a distinct lack of trust in any government introducing a new charge (it has been coined "traffic tax" in the media, which is disastrous) around what it does with the revenue and if it will reduce other taxes.  

This is why there is a need to talk about road reform more generally, and how congestion pricing can be offset by lower prices offpeak (by replacing registration fees and fuel taxes).  There is a need to bring the public along with how pricing can work, what it would replace and what revenue would be used for.  That requires a lot of effort.  To glibly talk about public acceptability in London (where hardly anyone actually drives to the CBD) or Stockholm (where similarly, most trips are not charged and revenue has been hypothecated for roads) is simply missing the point.  It is overwhelmingly obvious that the reason this policy hasn't gotten anywhere in Australia is because it is politically toxic and that is because it is toxic with the public.

It is that which the Grattan Institute needs to address, which is convincing the public that this would be good for them (and what "this" would look like).  I see little evidence of this, and the public backlash about the report is counterproductive.

What now?

More needs to be done, and it would be helpful to acknowledge that, of all cities, Auckland is more advanced in thinking than any Australian cities, not least because most recent reports indicate a central and local government are in some alignment about the need to act and that pricing is part of the mix.

Australia needs a conversation about pricing roads, which includes congestion pricing, which includes replacing registration and fuel tax, and most importantly discusses what is done with the money, how the roads are managed and paid for.   The Grattan Institute report contains some useful data and analysis, but a report that misconstrues the London scheme, that jumps to transplanting a cordon onto Sydney and Melbourne and ignores the national agenda of road reform falls well short.  My hope is that it doesn't undermine the whole argument by generating public opposition about the concept, by proposing options that are fundamentally flawed.

News Briefs - Australia, Brazil, China, India

Australia - Melbourne's car park congestion tax to stay despite low impacts

Public broadcaster ABC reports that the Victorian Government is likely to keep the tax on inner city car park spaces intended to reduce congestion, despite a study that claims it has had little effect on congestion.

Inner city parking taxes are often cited as low cost alternatives to implementing congestion pricing.  50,000 car park spaces in downtown Melbourne are subject to such a tax which is either A$650 (US$682) or A$910 (US$956) a year depending on the location.  The tax was introduced in 2006 and raises A$46 million (US$48 million) a year in revenue for the Victorian State Government.

A Monash University study indicated that part of the tax was being absorbed by car park operators and that there was minimal impact on congestion, it suggested that a better option would be a cordon based congestion charge, although motorist lobby group the RACV (Royal Automobile Club of Victoria) rejected that in a report in the Herald-Sun newspaper claiming motorists are already over taxed, preferring more money to be spent on transport projects.

Brazil - Expiring toll road concessions to be retendered

Nasdaq reports that according to the Estado de S Paulo newspaper, the Brazilian government has decided that when the first toll road concessions expire in 2015 it will re-tender them with the intention that rates of return should be lower.   It will also specify that tolls can only be charged after a certain proportion of improvements to highways have been undertaken by a concessionaire.

"To ensure that 5,700 kilometers of two-lane highways are expanded into four-lane highways by 2018, the government will stipulate that operators start charging tolls only after 10% of the lane-duplication process is already executed" ..."As part of a package to improve transportation infrastructure, the federal government plans to license 7,500 kilometers of highways to private operators for a 25-year period".

Concessionaires have expressed concern that environmental licencing imposes a cost on development that will slow down the period between when they start work on a project and can start charging tolls.

China - Beijing traffic plan includes concept of congestion pricing

Global Times reports that Beijing is to get congestion charging included in its latest traffic plan, although there are no details about what congestion pricing will look like for the city.  The article has comments from an academic who says Beijing also needs more rail and congestion is inevitable with fast growth, but also a local resident who uses public transport and supports the idea.  I wrote some time ago about potential options for congestion charging in Beijing.

India - Macquarie SBI buys into Indian concessionaire

According to NetIndian News Network, Australian/Indian joint venture Macquarie SBI Investment Fund (MSIF) (comprising Macquarie Capital Group and the State Bank of India) has announced it is investing US$150 million into Ashoka Concessions Ltd (ACL) of India, along with SBI Macquarie Infrastructure Trust.  It would appear likely that the reason for the MSIF (and SBI Macquarie Infrastructure Trust) is that Macquarie Capital is restricted from investing in its own right in Indian infrastructure because of foreign ownership limit laws (and the State Bank of India is keen to use foreign capital and expertise).

ACL is a 100% subsidiary of Ashoka Buildcon Ltd (ABL), an Indian engineering and construction firm.  ACL reportedly owns 7 concession toll roads comprising 3,018 lane km (630 length km). ABL itself has a portfolio of 12 Build Own Transfer (BOT) road projects in India (with 6 under construction) excluding its portfolio of footbridges.  Not all of these projects are tolled.

ACL's projects have a construction cost of US$1.5 billion.  78% of the traffic is commercial.  Average remaining concession period is 22.7 years.  This is MSIF's first road investment and will comprise 13% of MSIF's portfolio value.

Macquarie SBI published a presentation with more detail about the investment.

Melbourne Eastlink faces takeover crunch time

A casual observer of the Australian PPP toll road scene might be excused for being pessimistic about the finances of them. After all, more news is made of the ones that don’t do well than the many that do. Cross City Tunnel in Sydney, Clem 7 tunnel in Brisbane and the Eastlink toll road in Melbourne have all failed to meet expectations. However, in all of these cities are very successful privately owned toll roads as well, and even poorly performing ones have interest from investors with their eye on future growth.

M3 - Eastlink toll road

Melbourne’s Eastlink is one of those. Eastlink is 39km long and was fully completed in 2008. It forms a major north-south corridor through the eastern suburbs of the greater metropolitan area of Melbourne. It connects the untolled Eastern Freeway (a radial route from the north of the central city to the eastern suburbs to the untolled Monash Freeway (a radial route to the south east) and ends at the Mornington Peninsula Freeway. It is owned by ConnectEast, which has as its largest shareholder Australian Infrastructure Investment firm CP2 (35%), other major shareholders being Lazard Asset Management, RARE Infrastructure and the Commonwealth Bank of Australia.

CP2 is seeking to buy out ConnectEast in its entirety through its investment house Horizon Roads, which is a consortium of eight funds (includes the British Universities Superannuation Scheme, the National Pension Service of Korea, New Zealand Superannuation Fund and the Teachers Insurance and Annuity Association of America). The offer is worth A$2.17 billion (US$2.24 billion) or A$0.55 (US$0.57) per share. Given the shares for ConnectEast originally floated at A$1 a few years ago, and Deloitte has since claimed the offer is “fair and reasonable” within a range of A$0.51 and A$0.57 (US$0.53-US$0.59), there may be a reasonable chance of it proceeding as shareholders seek to get what they can. The shareprice jumped 20% when the offer was made, and is now around A$0.51. CP2 isn't supporting rival investor, Transurban, making a bid - which seems obvious, except that it has a minority stake in Transurban as well.

Presumably Horizon Roads thinks there are good prospects for the road. The Herald Sun reports  that the difference between the forecast and actual revenue per month is around A$242,237 (US$249,818). More recently, traffic and revenue have been climbing. ConnectEast reported a 11.9% rise in annual daily revenue for August 2011 compared to August 2010, with total revenue of A$661,525 (US$682,229) for the month. Traffic numbers were 194,555 in August 2011, which was a 8.5% increase on the previous year, indicating increasing yields.

The takeover offer has been given the clearance by the Foreign Investment Review Board, so shareholders will vote on it on 27 September 2011. It will be interesting to see if they think the offer is fair and reasonable. Some may see it as a chance to bailout of what looked like a bad investment, others may prefer to hang on, because the current investment climate may make the road look like a better prospect than some alternatives.

Whatever happens, it wont affect toll rates which under the concession can only be varied annually according to changes in the consumer price index. ConnectEast spokesman James Tonkin claims “EastLink will continue to have the cheapest per kilometre car tolls of any private tollway in Australia”. He’ll be hoping that persuades a few more motorists to consider the road is better value than they may otherwise think.

No congestion charge for Melbourne says new Deputy Premier

The recent State election in Victoria, Australia saw the Labour Party ousted for the centre-right Liberal/National coalition.  Melbourne has a well developed public transport system (focused on the central business district of course), and for some years there have been discussions about whether a congestion charge (focused on the rather easily defined central city) could help ease traffic congestion and fit in with the city's transport strategy.

The new state government seems to have made its views known through new Deputy Premier Peter Ryan who is reported by AAP as opposing a congestion charge for Melbourne.  

Unfortunately, he is making the common mistake around congestion charging in thinking that such charges automatically mean people need a public transport alternative.   The experience elsewhere is that a significant proportion of motorists do not change modes, but change travel times, consolidate trips or do not undertake trips at all.   Public transport providers should anticipate increased demand, but the real point is that public transport at peak times is also underpriced, so congestion charging is an opportunity for further reforms. 

Another option is not to introduce a new charge, but to rebalance existing charges so that it may cost less at other times.   This would make it advantageous to those driving off peak and would increase overall economic efficiency, and may be designed to be net revenue neutral.

Still, there remains some obvious attraction in introducing a congestion charge for downtown Melbourne predominantly because the city's rail, tram and bus network provides good standards of service into the city from the rest of the metropolis.   However, when most discuss such charges they think of cordon/area charges ala London, Stockholm, rather than considering more disaggregated charges that may vary by time and location.   Congestion charging for Melbourne cannot be blunt, cannot be all day and should not have the same charge on all roads approaching the central city.   Equally important is to consider whether congestion charging is better considered as a form of distance charging rather than single events (e.g. crossing a set point).

Congestion Charging advocated for Melbourne

Allan Fels, Dean of the Australia and New Zealand School of Government  has said in the Australian newspaper that congestion charging would be one of the best measures to improve economic efficiency in the state of Victoria, Australia.

Victoria has already investigated traffic congestion and congestion charging before, in a comprehensive report by the Victorian Competition and Efficiency Commission.   The response of the Victorian government was to embrace many of the proposals around new infrastructure and public transport services, but to sidestep urban congestion pricing.  Given Melbourne's extensive CBD centred public transport network, pricing road access to the CBD (which already exists on two tolled corridors) does not appear unreasonable, but is politically difficult.  Although parking is effectively priced and rationed in a way that forms a second-best proxy to congestion charging.  Yet urban CBD charging alone would only address some of Melbourne's congestion problems given the very large area of the metropolis and high usage of cars for trips that do not focus on the CBD.   This suggests that wider reform of road charging policy (replacing fuel taxes and ownership taxes with distance based charging) may have more applicability.

Yet the mere fact it is being discussed in Australia is progress on its own.   Simply building roads and building public transport networks in themselves does not address the fundamental problem with not efficiently pricing roads - when demand exceeds supply, queuing eventuates.

It follows on from reports earlier this year that Infrastructure Australia is supporting congestion charging to make cities more sustainable, and to raise revenue to improve transport networks.